Oncotype DX Genomic Prostate Score

Clinical Evidence for the Oncotype DX Genomic Prostate Score

Developed, validated, and studied in over 4,000 patients1-9



4,000
Patients1-9

Oncotype DX GPS was specifically designed to guide treatment decisions in patients with clinically low-risk prostate cancer. Over 4,000 patients have been included in development studies, clinical validation studies, and clinical utility studies.1-9

Validation studies

The validation studies for the Oncotype DX Genomic Prostate Score (GPS) were conducted in a prospective fashion (using retrospective cohorts) in clinically low-risk patients from large, contemporary databases:

  • Kaiser Permanente, Northern California (KPNC) (N=259).1
  • University of California, San Francisco (UCSF) (N=395).2
  • Center for Prostate Disease Research (CPDR) (N=402).3

A meta-analysis combining the UCSF and CPDR cohorts (N=732)9 validated the association between Oncotype DX GPS and refined estimation of the likelihood of favorable pathology.

Findings summary

The Oncotype DX GPS assay has been proven to be an independent predictor of both clinical risk and tumor aggressiveness:

  • Clinical risk: Validated as an independent predictor of both prostate cancer death and metastasis at 10 years1
  • Tumor aggressiveness: Validated as an independent predictor of adverse pathology, the most immediate and actionable endpoint in patient management2,10,11

KPNC cohort1

GPS Per 20 Units Multivariable*
Odds Ratio P Value
Prediction of prostate cancer death 2.7 <0.001
Prediction of metastasis 2.3 <0.001
*Adjusted for NCCN clinical risk criteria

UCSF cohort2

GPS Per 20 Units Multivariable*
Odds Ratio P Value
Prediction of high-grade disease 2.3 <0.001
Prediction of non-organ-confined disease 2.0 0.002
Adverse pathology† 1.9 0.001
*Adjusted for NCCN clinical risk criteria
†Defined as high-grade (primary Gleason pattern 4 or any pattern 5) and/or non-organ-confined prostate cancer

The Oncotype DX GPS independently predicts adverse pathology after adjusting for clinical risk factors.

CPDR cohort3

GPS Per 20 Units Multivariable*
Odds Ratio P Value
Adverse pathology† 3.3 <0.001
Biochemical reccurence Hazard Ratio P Value
2.7 <0.001
 
GPS Per 20 Units Univariable*
Odds Ratio P Value
Metastatic recurrence‡ 3.8 <0.001
*Adjusted for NCCN clinical risk criteria
†Defined as high-grade (primary Gleason pattern 4 or any pattern 5) and/or non-organ-confined prostate cancer
‡Small number of events preclude multivariable analysis

The Oncotype DX GPS independently predicts adverse pathology and biochemical recurrence after adjusting for clinical risk factors.

Meta-analysis cohort9: GPS Adds Predictive Value for the LFP to Different Clinical Risk Assessment Tools

UCSF CPDR Meta-analysis
P Value* P Value P Value
GPS and CAPRA Score vs CAPRA Score Alone <0.001 <0.001 <0.001
GPS and NCCN Risk Group vs NCCN Risk Group Alone 0.002 <0.001 <0.001
GPS and AUA/EAU Risk Group vs AUA/EAU Risk Group Alone 0.002 <0.001 <0.001
*P-value from likelihood ratio test with degree of freedom of one for each individual study and two for meta-analysis

The Oncotype DX GPS independently predicts the likelihood of favorable pathology regardless of clinical risk tool used.

Patient population9: Clinical Characteristics

Meta-Analysis Cohort
(N = 732)
Age at Diagnosis Median 60 years
Range 38-77 years
Race Caucasian 614 (84%)
African American 81 (11%)
Other 37 (5%)
PSA 0 to < 4 159 (22%)
4 to < 10 493 (67%)
10 to 20 80 (11%)
Clinical T-Stage T1 462 (63%)
T2 270 (37%)
Biopsy Gleason Score ≤ 3+3 559 (76%)
3+4 173 (24%)
NCCN® Risk Group Very Low 78 (11%)
Low 375 (52%)
Intermediate 263 (37%)

By combining the two validation datasets, the meta-analysis cohort was more reflective of the US prostate cancer population

REFERENCES

1. Van Den Eeden et al. Eur Urol2017.
2. Klein et al. Eur Urol. 2014.
3. Cullen et al. Eur Urol. 2014.
4. Badani et al. Urol Pract. 2015.
5. Data on file. Genomic Health, Inc.
6. Dall’Era et al. Urol Pract. 2015.
7. Knezevic et al. BMC Genomics. 2013.
8. Hamdy et al. N Engl J Med. 2016.
9. Brand et al. Urology. 2016.
10. Kozminski et al. Urol Oncol. 2016.
11. Eggener et al. J Urol. 2011.

a. DOI: DOI: 10.1200/JCO.2016.34.15_suppl.e16611 
b. DOI: 10.1016/j.juro.2017.02.637
c. DOI: h10.1016/j.eururo.2014.05.004
d. DOI: 10.1016/j.eururo.2014.11.030

 

Clinical Utility:
VA Medical Centersa

The impact of Oncotype DX GPS on prostate cancer management strategies was studied in 6 VAMCs.

Read Full Text

Clinical Validation:
KPNCb

Kaiser Permanente of Northern California assessed the association between GPS and two hard endpoints: metastasis and prostate cancer death in a cohort of 279 men.

Read Full Article

Clinical Validation:
UCSFc

The first validation study for Oncotype DX GPS analyzed needle biopsies from consecutive patients at the University of California San Francisco.

Read Full Text

Clinical Validation:
CPDRd

A confirmatory validation study at the Center for Prostate Disease Research included a relatively high proportion of African-American patients.

Read Full Text

Stay in Touch

Have a specific question? Want to keep up with our latest news? We're here for you.

Contact Us
Making cancer care smarter.™